Substituted Nitroquinolines Immobilized in Multiwalled Carbon Nanotubes: An Unconventional Voltammetric Experiment

Multiwalled carbon nanotube-glassy carbon electrodes (MWCNT-GCEs) were prepared and used to immobilize 5-nitroquinoline (5-NQ), 6-nitroquinoline (6-NQ) and 8-nitroquinoline (8-NQ). Cyclic voltammetry was used to study the conventional voltammetric behavior of these NQs dissolved in an aqueous alcoholic solution and compare them with the unconventional voltammetric behavior observed when the NQs were immobilized in a porous layer of MWCNTs. Increased currents and shift to lower overpotential can be explained by the change of the mass transport regime from semi-inﬁnite diffusion to thin layer diffusion. Reduction peak potentials depended on the position of the nitro group (-NO 2 ) in the quinoline structure and the test cell pH. Reduction peak potential of -NO 2 was dependent on the mesomeric effect of the resonant structures and the inductive effect of the N heteroatom. ArNO/ArNHOH redox couple was electrogenerated in situ and immobilized on a nanostructured electrode network for all the studied NQs. The redox couple generated from 5-NQ was the most stable with 50% remaining at the 15th cycle. Even though the current intensity observed for 6-NQ was the largest, the stability of 6-NQ was 35% at the same voltammetric cycle. The lowest stability was observed for 8-NQ with only 25% remaining at the 15th cycle.

The use of carbon nanotubes has dramatically changed the molecular architecture of conventional electrodes, producing an enormous variety of new modified electrodes. Modified electrodes can achieve objectives that are impossible to achieve with conventional electrodes and considerably expand the frontiers of electrochemical science. These modified electrodes have applications in many areas of electrochemistry, such as electrocatalysis, sensing and biosensing, corrosion, batteries, clinical diagnosis, and point-of-care devices.
The use of multiwalled carbon nanotubes (MWCNTs) to modify conventional electrodes has grown over time. Many studies have been published about glassy carbon electrodes (GCEs) or screen printed carbon electrodes (SPCEs) modified with multiwalled carbon nanotubes (GCE-MWCNTs), and the nanomaterials have been modified through chemical reactions. [1][2][3][4] In recent years, we have developed a simple procedure to modify MWCNTs with nitroaromatic compounds such as nitrobenzene, nitrobenzene-substituted 1,4dihydropyridine derivatives, 2,7-dinitro-9-fluorenone, 3-nitrobenzoic acid, 3,5-dinitrobenzoic acid and 1,3,5-trinitrobenzene. [5][6][7][8] The procedure involves dipping the open circuit of an electrode modified with MWCNTs in a solution containing the nitroaromatic compound. Using this procedure, the nitroaromatic compound is trapped in the three-dimensional MWCNT network with sufficient stability to be subjected to an in situ electrochemical experiment, allowing applications to electrocatalytic processes and unconventional methods of performing voltammetry.
Consequently, by potential cycling at a modified electrode, an ArNO/ArNHOH redox couple was electrogenerated in situ and immobilized on a nanostructured electrode network. Surprisingly, this redox couple was capable of acting as a mediator for the electrocatalysis of NADH. 9 This new method to modify MWCNT electrodes with electroactive compounds is an unconventional way to perform voltammetry in which a non-aqueous solvent is first used to trap the organic electroactive species in a porous surface layer, i.e., MWCNTs, and the modified electrode is then transferred to an aqueous buffer to conduct voltammetry measurements. 10 A rather similar strategy was followed by P. Fanjul-Bolado et al., with the adsorption of leucoindigo on CPE and exchanging the electrolyte solution for detection. 11 In this work, we will test this unconventional voltammetric approach using a comprehensive nitroaromatic compound that can be trapped within the three-dimensional network of MWCNTs. Obviously, not all nitroaromatic compounds behave identically, and our current challenge is to find nitroaromatic compounds that have sufficiently stable interactions with MWCNTs. The objective is to find general parameters that describe nitro compounds of interest to modify MWCNTs. According to our previous studies and state-of-the-art research, we selected some characteristics to find new nitroaromatic compounds that are suitable for adequately immobilizing a threedimensional array of MWCNTs. First, the nitroaromatic compound must be poorly soluble in an aqueous buffer, and second, the nitroaromatic compound must be highly conjugated and contain π-delocalized electrons that facilitate interactions with MWCNTs.
In this manuscript, we chose molecules with highly delocalized electrons, such as those in the quinolenic ring, with a nitro group in different positions. In fact, we selected 5-nitroquinoline (5-NQ), 6-nitroquinoline (6-NQ) and 8-nitroquinoline 8-NQ ( Figure 1). Nitroquinolines are dangerous genotoxic substances with potential carcinogenic and mutagenic effects, 12,13 and quinoline is also a detrimental substance. 14 Nitroquinolines are formed during incomplete combustion of both gasoline and diesel fuel and during the biomass pyrolysis process. 15 The electrochemical reduction of nitroquinolines has been studied with mercury electrodes, 16 silver solid amalgam electrodes, 17 silver solid electrodes, 18 carbon-film electrodes 19 and carbon-fiberrod electrodes. 20 However, to date, the electrochemical behavior of nitroquinolines on carbon nanotubes had not been explored. The aim of this study was to find the optimum conditions to carry out unconventional voltammetry with electroactive species (NQs) trapped in an electrode instead of dissolved in solution and to determine the effect of the nitro-substitution position on the electrochemical behavior of the trapped species.

Experimental
Reagents and apparatus.-5-Nitroquinoline (5-NQ), 6nitroquinoline (6-NQ) and 8-nitroquinoline (8-NQ) were purchased from Sigma-Aldrich. All other reagents were of analytical grade. MWCNTs (10 nm in diameter and 1.5 μm in length) were purchased from Dropsens S. L, Spain. Stock solutions of NQs were prepared in ethanol. Phosphate and Britton-Robinson buffers were used; pH adjustment was performed using a WTW PMX 3000 pH/ion meter (for basic pH adjustments) and a WTW 537 pH microprocessor pH meter (for acidic pH adjustments). Electrochemical measurements (cyclic voltammetry, CV) were carried out with a voltammetric analyzer (BASi 100, BASi analytical Instruments, USA). A CHI 650C potentiostat (CH Instruments Inc., USA) and a multichannel potentiostat (MultiEmStat, PalmSense, The Netherlands) were also used. The measurements were performed with a conventional 3electrode system: a GCE with a 3 mm diameter (model CHI104, CH Instruments), Ag/AgCl/NaCl (BASi MF-2052) and a Pt wire were used as the working, reference and counter electrodes, respectively.

Preparation of the modified electrodes.-The
GCEs were washed and polished with 0.3 and 0.05 μM alumina and thoroughly rinsed with water. The MWCNTs were dispersed following the procedure described in Ref. 9. Briefly, a dispersion (3 mg mL −1 ) was prepared in 1,3-dioxolane by sonicating the mixture three times for 5 min. Then, 5 μL of the MWCNT dispersion were deposited on the electrode surface, and the solvent was subsequently evaporated. The resulting modified electrode is designated as GCE-MWCNT.

Entrapment of NQ in the GCE-MWCNT network.-The GCE-
MWCNT was immersed in a solution of 0.1 mM of the nitro compound in ethanol for a given time, i.e., the accumulation time (t ac ). After this, the electrode was washed with water to remove traces of solvent and any possible excess of nitro compound not trapped on the electrode. After washing, the electrode containing entrapped NQs was placed in an aqueous buffer for the voltammetric experiments.
Voltammetric measurements.-For unconventional voltammetry, the modified electrodes containing electroactive NQ were immersed in a cell containing only the aqueous buffer without a nitro compound. Conventional voltammetry was performed using a GCE immersed in an electrochemical cell containing electroactive NQ dissolved in an aqueous buffer-ethanol solution. All electrochemical measurements were performed after purging the cell solution with N 2 for 5 min. Experiments were carried out at room temperature.

Results and Discussion
Electrochemical behavior of substituted NQs entrapped in MWCNT-GCE.-First, the conventional cyclic voltammograms of NQs in solution were obtained using a GCE ( Figure 2A). These NQs exhibited the typical electrochemical behavior of nitroaromatic compounds in buffer solutions; 21 i.e., for a potential scan starting at −0.3 V and scanning in the negative direction, an irreversible reduction peak appears (peak N1) with a peak potential (E p ) that depends on the nitrosubstituent position on the NQ. Thus, E p values of −0.52, −0.58 and −0.47 V were observed for 5-, 6-and 8-NQ, respectively. This peak corresponds to ArNO 2 reduction, as shown in Equation 1, producing the reduced hydroxylamine derivative (ArNHOH).
In the reverse scan, oxidation of the hydroxylamine derivative is observed at approximately 0.0 V (peak N2a), resulting in the derivative ArNO. Finally, a second negative sweep reduces the ArNO derivative to ArNHOH (N2b peak). Thus, peaks N2a and N2b form a quasireversible redox couple, as described by the well-known Equation 2.
As previously established by our research group, some nitroaromatic compounds can be entrapped in the three-dimensional network of MWCNTs, and unconventional voltammetry can be performed. 10  Figure 3D shows the stability (expressed as percentage of gradual change in the current intensity) as a function of the number of cycles, and 5-NQ was the most stable with 50% remaining at the 15th cycle. Even though the current intensity observed for 6-NQ was the largest (Figure 3B), the stability of 6-NQ was 35% at the same voltammetric cycle. The lowest stability was observed for 8-NQ with only 25% remaining at the 15th cycle.

Influence of the accumulation time and the concentration of NQ on entrapment.-To study the effect of the accumulation time and NQ
concentration on entrapment, the procedure described in Experimental sections of this manuscript was used. Voltammograms were recorded at 0.1 V/s after an accumulation time of 10 s. The results showed that in the studied concentration ranges, the current intensity of the nitro group (N1) increased as the concentration of NQ increased for all the compounds. In Figure 4 we show the effect for the 6-NQ derivative but all the studied NQs follow a similar behavior. The dependence of the peak current on the accumulation time was evaluated using solutions of NQ at different concentrations to modify the MWCNT network. As shown in the insert of Figure 4, for 6-NQ, adsorption was practically instantaneous and reached a virtually constant peak intensity value after the 1 s. for concentration of 0.1 mM. In the case of the lowest concentrations the maximum was reached at somewhat higher times.

Influence of pH.-
The effect of the pH on the reduction peak N1 and the reversible couple (N2a/N2b) peaks was also studied by CV. Voltammograms of GCE-MWCNTs modified with NQs and immersed in a buffer with supporting electrolyte were recorded at pH 2-11. Figure 5 shows the voltammograms obtained at pH values of   2.0, 7.0 and 11.0. Figure 6 depicts the influence of the pH on the reduction peak N1 over the entire pH range. In general, all the peaks shifted toward more negative potentials as the pH of the supporting electrolyte buffer increased. Previous studies on the effect of pH on the reduction peak potential of nitroaromatic compounds have shown this effect. 22 However, it is important to emphasize that in the present study, no nitroaromatic compounds were present in solution because they were completely immobilized in the MWCNT network. Figure 6 shows that the dependence of E p on pH is higher at an acidic pH than that at a basic pH. At pH 2, 5-NQ and 8-NQ showed nearly the same N1 peak potential, while at basic pH values (9 to 11), 5-NQ and 6-NQ had almost the same peak potential. At acidic pH values, the major species contains a protonated quinoline nitrogen. A change in the slope of E p vs pH upon the reduction of the nitro group (N1 peak) was observed for all cases. The linear range was from pH 2.0 to pH 8.0 for 5-NQ and 6-NQ and pH 2.0 to pH 7.0 for 8-NQ. The slope values (m) are summarized in Table I. The slopes for all observed voltammetric peaks (N1, N2a and N2b) were very close to the theoretical value of an e − /H + ratio of 1:1 (slopes near −0.059 V, see Table I). These results are consistent with the stoichiometric ratio of transferred protons and electrons in Equations 1 and 2. However, changes in the slopes of the N1 peaks for the three compounds suggest a change in the reaction mechanism. Slopes values between −0.024 and −0.034 V were obtained in the pH range between 8.0 and 11.0, and these values correspond to an e − / H + ratio of 2:1 for the rate-determining step of this reaction.
Based on the possible effect of pH on the voltammetric signals of 5-NQ observed in previous work, [16][17][18][19] the -NO 2 group reduction of the 5-NQ peak (N1 signal) is split into two signals at basic pH values or in aprotic media; i.e., an initial 1-electron reduction step to form the ArNO 2 •− radical (Equation 3), followed by a 3-electron reduction step to form the hydroxylamine derivative (Equation 4).  When GCE-MWCNTs is used, this split was not observed. Previous studies have shown that the formation of the species ArNO 2 •− is favored on a solid electrode because the electron transfer rate is low. 23,24 MWCNTs are nanomaterials with good conductive properties that facilitate electron transfer, which could be the reason why the main ArNO 2 reduction signal was not divided into two signals at alkaline pH values (Equations 3 and 4). Nevertheless, changes in the oxidation-reduction mechanism of the redox couple with pH when working with NQ derivatives trapped on this specific surface cannot be ruled out.
Effect of the nitro group position on the peak potential.-As shown in Figures 5 and 6, 8-NQ was clearly reduced at the lowest potential, while 6-NQ was reduced at the highest potential. The potential values of the N1 peak are directly related to the energy required to reduce the -NO 2 group. The experimentally obtained data can be interpreted by taking into account the position of this group on the molecule. 22,25 Some effects can explain these results. First, consider the aromatic quinoline skeleton of these three compounds. Structures such as pyridine or quinoline are deficient-π systems, and the N heteroatom is a strong electron-withdrawing group. N has a nonbonding electron pair, but the pair is perpendicular to the electron π-cloud of the entire molecule. Thus, this pair of electrons does not participate in the aromaticity, resulting in this kind of molecule acting as a strong base. Resonance structures of a quinoline skeleton (Figure 7) show that the 5-position results in deficient electron density (structure IV, Figure 7). Since the N atom in the -NO 2 group also has a positive charge, the close proximity of both positive charges generates instability in the molecule, favoring the reduction of this group. This was observed for compound 5-NQ, in which the nitro group (N1 peak) was more easily reduced than that in 6-NQ.
However, these resonance structures do not explain the lower E p of 8-NQ compared with that of the two other compounds. Since the distance between the 8-position of the quinoline skeleton and the electronegative N heteroatom is only two carbon bonds, the inductive effect of the heteroatom predominates over the mesomeric effect (resonance structures). Thus, the electronegative N atom attracts the electron density from adjacent atoms, 26 resulting in the carbon atom in position 8 having a positive charge density. This effect, in addition to the positive charge of the N atom in the -NO 2 group, causes the molecule to be unstable, and reduction of the -NO 2 group is favored.
Thus, a mesomeric effect of electron withdrawing from the N heteroatom at the 5-position in the quinoline skeleton is obtained, resulting in the nitro group reduction E p of 5-NQ changing to a less negative potential close to the E p of the 8-NQ nitro group reduction and separating the reduction peak potentials of 5-NQ and 6-NQ. Since the 8-NQ reduction is mainly affected by the inductive effect of the N heteroatom in the molecule, this signal appears at the lowest potential.
Another important factor is the nature of the quinoline moiety of the molecule, which is protonated or not depending on the pH. 27 At acidic pH values, the N heteroatom in the molecule is protonated, increasing both mesomeric and inductive effects from this heteroatom.
At a neutral pH, the protonated and deprotonated species are equal, and the mesomeric effect discussed above begins to decrease. Finally, at basic pH values, the N heteroatom is not likely to be protonated, and the mesomeric effect is almost completely eliminated. The N1 peak is almost the same for 5-NQ and 6-NQ, and the 8-NQ reduction E p is a less negative potential than that of the other compounds. These results are consistent with the inductive and mesomeric effect discussions mentioned above.
Surprisingly, this behavior shows that the molecular electronic effects that manifest for species in solution in conventional voltammetry are equivalent to those that manifest for species trapped in the MWCNT network in unconventional voltammetric experiments.

Conclusions
All the NQs studied in this work can be immobilized or trapped in a porous layer of MWCNTs through weak interactions (e.g., hydrogen bonding, π-π stacking, electrostatic forces, Van der Waals forces and hydrophobic interactions) without the formation of covalent bonds. This form of immobilization is particularly attractive because it allows chemical functionality to be attached while preserving the sp2 structure of the nanotubes, leaving the electronic structure unchanged.
Using the above approach, it is possible to trap NQ species in a porous network of MWCNTs on an electrode to perform unconventional voltammetry, which prioritizes thin-layer-type mass transport and results in larger peak currents and smaller concentration overpotentials. The same conclusion was reached by different authors under very different circumstances. [28][29][30][31][32][33] We have shown that the position of the nitro group in an aromatic structure (quinoline) results in different voltammetric responses and adsorptive properties for 5-, 6-and 8-NQ trapped in a MWCNT-GCE network. The inductive effect of the N heteroatom in the molecule reduces the reduction potential of the nitro group in position 8. The mesomeric effect favors the reduction of this group when it is in position 5 at acidic pH values and a neutral pH and decreases as the pH increases.
An ArNO/ArNHOH redox couple was electrogenerated in situ and immobilized on the nanostructured electrode network for all the studied NQs. The redox pair generated from the NQs promoters and their adequate stability represent an important result that can lead to the generation of interfaces with improved electrocatalytic characteristics, as has been demonstrated for other nitro compounds. 3,9